Samavedam Santhi and Srinivasan Sundarrajan

ФАЗОВЫЕ ПРЕВРАЩЕНИЯ

PACS numbers: 07.05.Tp, 61.72.Qq, 64.70.dg, 81.05.Bx, 81.20.Hy, 81.30.Fb

Shrinkage Estimation of Cast Al–Si Alloys Through Process

Simulation

Samavedam Santhi and Srinivasan Sundarrajan*

Mahatma Gandhi Institute of Technology, 500075 Hyderabad, India *National Institute of Technology, 620015 Trichy, India

With the advent of various casting process simulation techniques, it is possi-ble to derive theoretically casting parameters, which can be further validated

through experiments. In this paper, detailed simulation activities are carried

out to determine the shrinkage characteristics of two Al–Si alloys (US 413

and US A356) followed by validation through experimental studies. Produc-tion of defect-free castings requires good understanding of shrinkage charac-teristics. The influence of various process parameters on these characteris-tics determines the casting quality. Based on this, an extensive study is un-dertaken to comprehend the influence of various parameters.

Keywords: shrinkage, cast aluminium alloys, casting mould, process simula-tion, solid model, design of experiments.

З появою різних метод моделювання процесу лиття з’явилася можливість

теоретично оцінити параметри лиття, які можуть бути додатково переві-рені за допомогою експериментів. В даній роботі було проведено детальне

моделювання для визначення характеристик усадки двох стопів Al–Si (US 413 і US A356) з подальшою перевіркою за допомогою експерименту. Виробництво виливків без дефектів потребує доброго розуміння характе-ристик усадки. Вплив різних параметрів процесу на ці характеристики

визначає якість виливки. Виходячи з цього, було проведено цілу низку

досліджень, щоб визначити вплив різних параметрів.

Ключові слова: усадка стопу, ливарні алюмінійові стопи, ливарна форма,

Corresponding author: Samavedam Santhi E-mail: [email protected] Please cite this article as: Samavedam Santhi and Srinivasan Sundarrajan, Shrinkage

Estimation of Cast Al–Si Alloys through Process Simulation, Metallofiz. Noveishie

Tekhnol., 39, No. 7: 959–981 (2017), DOI: 10.15407/mfint.39.07.0959.

Ìåòàëëîôèç. íîâåéøèå òåõíîë. / Metallofiz. Noveishie Tekhnol. 2017, т. 39, № 7, сс. 959–981 / DOI: 10.15407/mfint.39.07.0959 Îттиски доступнû непосредственно от издателя Ôотокопирование разрешено только в соответствии с лицензией

2017 ÈÌÔ (Èнститут металлофизики им. Ã. В. Êурдюмова ÍÀÍ Óкраинû)

Íапечатано в Óкраине.

959

mailto:[email protected]

960 Samavedam SANTHI and Srinivasan SUNDARRAJAN

симуляція процесу, модель твердого тіла, проєктування експериментів.

С появлением различнûх методов моделирования процесса литья появи-лась возможность теоретически оценить параметрû литья, которûе могут

бûть дополнительно проверенû с помощью экспериментов. В данной ра-боте бûло проведено детальное моделирование для определения характе-ристик усадки двух сплавов Al–Si (US 413 и US A356) с последующей про-веркой с помощью эксперимента. Производство отливок без дефектов тре-бует хорошего понимания характеристик усадки. Влияние различнûх

параметров процесса на эти характеристики определяет качество отлив-ки. Èсходя из этого, бûло проведено обширное исследование, чтобû опре-делить влияние различнûх параметров.

Ключевые слова: усадка сплава, литûе алюминиевûе сплавû, литейная

форма, моделирование процессов, модель твёрдого тела, проектирование

экспериментов.

(Received June 29, 2017)

1. INTRODUCTION

Solidification of castings involves modification of phase with libera-tion of latent heat from a moving liquid-solid boundary to mould and

then to atmosphere [1]. Casting process simulation programmes are

time and temperature dependent and more attention is being focused

on systematic aspects in production of castings such as studies of the

heat transfer mechanisms, by which castings solidify, and the rate, at

which solidification and shrinkage takes place [2]. Computer simulation of casting takes into account the qualitative

and quantitative consideration of the process parameters, boundary

conditions and their influence on quality of casting. Process parame-ters and boundary conditions include casting type, mould material,

chemical composition of alloy considered, physical properties and in-terfacial heat transfer coefficient between the existing boundaries. The temperature history of all sites inside the casting are attained

by plotting the progress of solidification fronts (isothermal contours)

at different instants of time, and by identifying the last freezing re-gions/sites [3]. The numerical simulation of casting solidification pro-cedure has been done through Finite Difference Method (FDM). The cast Al–Si alloys are the true workhorses of the aluminium alloy

casting industry because of their excellent casting characteristics and

good strength [4]. Shrinkage is the most significant discontinuity in

castings, and forms due to the contraction of liquid metal, transfor-mation of liquid to the solid and contraction in the solid state [5]. This

shrinkage results in the form of voids like micro- and macroshrinkage

during solidification. Macroporosity or macroshrinkage occurs due to

entrapped liquid metal surrounded by solid. Y. Li states [6] that the

SHRINKAGE ESTIMATION OF CAST Al–Si ALLOYS THROUGH SIMULATION 961

tendency for formation of shrinkage porosity is related to liquid to sol-id volume fraction at the time of final solidification and solidification

temperature range of alloy. Shape and relevant wall thickness of a

casting are very important casting parameters that influence the

shrinkage porosity.

2. SIMULATION PROCEDURE

Efficient design of experiments is an efficient approach for improving

a process to quickly obtain meaningful results and draw conclusions

about how factors or process parameters interact [7, 8] when more than

one factor is changing at a time. An orthogonal array would mean a

balanced design with equal weight age to each factor [9, 10]. MINITAB

software was used for experimental design [10].

2.1. Selection of Process Parameters

The parameters, which influence the shrinkage, are pouring tempera-ture, casting shape, use of chills, alloy composition, type of mould

sand, presence of mould coat and pouring time [11]. Out of the above parameters, which influence the shrinkage, alloy

composition, bottom chill, shape of the casting, mould coat, and pour-ing temperature, were considered as the process parameters for the

present investigation. Alloy composition affects the shrinkage charac-teristics and mechanical properties of the cast product. To study the

influence of alloy composition, two alloys with chemical compositions

given in Table 1 (US 413 and US A356) were used.

TABLE 1. Chemical composition (wt.%).

Element US 413 US A356

Silicon Iron

Copper Manganese Magnesium

Nickel Zinc Lead Tin

Titanium Others (each) Others (total) Aluminium

11.5 0.40 0.15 0.55 0.10 0.10 0.15 0.10 —

0.20 0.05 0.15

Balance

6.90 0.45 0.20 0.35 0.55 0.15 0.15 0.15 0.05 0.15 0.05 0.15

Balance


Rectangle, cube and cylinder were the three basic industrial casting

shapes considered for the present study. The rate of heat exchange de-pends on thermo-physical properties of the liquid metal, casting

mould, wall thickness of casting and its shape. The dimensional details

of the casting are provided in Table 2. Mould coatings provide smooth surfaces, improve the quality and

reduce the mould erosion [1, 12]. Pouring temperature [13] influences

fluidity, porosity, strength and structure of the casting. Pouring tem-perature T (T is melting point +75°C) and T + 50°C of superheat was

considered. The bottom chill showed significant difference in the cast-ing characteristics, and directional solidification can help in eliminat-ing shrinkage porosity defects [14]. A mild steel (MS) chill was consid-ered for the present study. Table 3 shows the details of the factors and their levels (variables)

for the present study. An orthogonal array L36 (24 3

1) was used with

five factors (one process parameter with three levels and four process

parameters with two levels) and 36 runs for shrinkage simulation in-vestigation, as shown in Table 4.

TABLE 2. Dimensional details of the casting shapes to determine shrinkage

characteristics.

Sl. No. Shape Dimension,

mm Pouring position

Surface area, cm2

Volume, cm3

1 Rectangle 115×100×48

436.40 552.000

2 Cube 82×82×82

403.44 551.368

3 Cylinder ∅90×90

381.51 572.265

TABLE 3. Factors and their levels for shrinkage simulation studies.

Factor 1 Alloy

Factor 2 Chill

Factor 3 Pouring

temperature, °C

Factor 4 Mould coat

Factor 5 Casting

mould

Level 1 US A356 Mild

Steel (MS) T + 50 Graphite (GC) Cylinder

Level 2 US 413 No T No coating (NC) Cube

Level 3 — — — — Rectangle


3. CASTING PROCESS SIMULATION

3.1. Solid Model

Simulation studies on shrinkage characteristics of aluminium alloy

have been conducted using the Virtual casting commercial software

TABLE 4. L36 (24 3

1) orthogonal array for the simulation studies.

Simulation

run order Alloy Chill Pouring

temperature, °C Mould

coat Casting

mould

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413

MS MS MS MS MS MS MS MS MS No No No No No No No No No MS MS MS MS MS MS MS MS MS No No No No No No No No No

T + 50 T + 50 T + 50 T + 50 T + 50 T + 50

T T T

T + 50 T + 50 T + 50

T T T T T T T T T T T T

T + 50 T + 50 T + 50

T T T

T + 50 T + 50 T + 50 T + 50 T + 50 T + 50

GC GC GC NC NC NC GC GC NC NC NC GC GC GC GC NC NC NC NC NC NC GC GC GC NC NC NC GC GC GC NC NC NC GC GC GC

Cylinder Cube

Rectangle Cylinder

Cube Rectangle Cylinder










Cube Rectangle


[15]. The solid models for shrinkage simulations were created using

the SolidWorks software. The solid models for shrinkage simulation are given in Fig. 1, along

with their respective chills where applicable.

3.2. Simulation Studies

Casting process simulations were carried out using Virtual Casting

software [15], which is a program for the simulation of the solidifica-tion process of industrial castings using FDM. Virtual casting software creates a virtual environment for casting

solidification, predicting and analysing the occurrence of shrinkage

defects. The input data for the casting process simulation are solid

model of the casting, material properties and boundary conditions. Virtual Casting consists of three major processes: Pre-processing,

Solving the governing equations and Post-processing or visualization

of results. The simulation output/results were displayed as contour

plots of temperature, porosity, solid fraction (Fs), solidification time

and cooling rate. For accurate results, the process simulation requires boundary con-ditions. The boundary condition values described at the beginning of

the process were the thermal data of metal and casting mould, condi-

Fig. 1. Solid models for shrinkage studies: a—shrinkage, b—respective chills.


tions of heat exchange between the casting and individual parts of the

mould and between the mould and its surroundings. The interfacial heat transfer coefficient is the rate of heat loss

through the metal/mould interfaces, which influence the casting char-acteristics [16]. However, interfacial heat transfer coefficient is not a

simple material property and is dependent upon chemical, physical in-terfacial conditions, mould and casting material properties and casting

geometry. The selection of interfacial heat transfer coefficient values,

as well as boundary conditions at the metal/mould interface, affects

the accuracy of the simulations. In the present investigations, interfa-cial heat transfer coefficient values for conformity of computer simu-lation and the experimental measurements were consistent. The ther-mo-physical properties of US A356, US 413 alloys, silica sand and mild

steel chill for the casting process simulation [16–18] are given in Ta-ble 5. The solid model of shrinkage characteristic, which is imported in the

STL-file format (Fig. 2) as the solution domain, divides into small fi-nite cells of casting and mould with a material identification. Bounda-ry conditions were assigned at all material interfaces metal, mould and

mould coat. The critical liquid fraction (a parameter used is the solvus in the

governing equations for calculation of porosity) is the value of liquid

fraction, below which a neighbouring cell does not feed to pay off for

shrinkage. During solidification at a particular location, liquid is

pulled in from neighbouring locations to pay off for solidification

shrinkage. If the neighbouring location has a high percentage of solid,

liquid cannot flow through it. The exit solid fraction is the total vol-ume percent of solid, at which the simulation exits.

TABLE 5. Physical constants and properties of US A356, US 413 alloys, silica

sand and mild steel chill for casting simulation.

Sl. No. Parameter US A356 US 413 Sand MS Chill

1 2 3 4 5 6 7

Melting point, K Thermal conductivity,

W/(mm⋅K) Density of solid, g/cm3 Liquids temperature, K

Freezing range, K Latent heat of fusion, J/kg

Specific heat, J/(kg⋅K)

934 0.159249

2.68496

886 400

388442 962.944

855 0.121338

2.65772

847 303.75 389112 1170

— 90.27×10−5

1.5219

— — —

1076.0076

— 4.5×10−2

7.84 — — —

460.548

8

Heat Transfer Coefficient, HTC, W/(m2⋅K)

Metal–mould Metal–coating mould

35×10−4 15×10−4

25×10−4 12×10−4

— —

— —


For example, a value of 1 for the exit solid fraction means that the

simulation will be done until the casting is 100% solidified. The simu-lation results for shrinkage are represented as contour plots of the lo-cation and magnitude of porosity formed during the solidification at

particular location from the bottom of the casting for simulation run

order 30 in Fig. 3. In Figures 3 and 4, the top portion shows the magnitude of porosity

formed during the solidification and the bottom portion depicts the lo-cation of the porosity from the bottom of the casting.

Fig. 2. STL-file and boundary conditions.

Fig. 3. Typical contour plots showing the porosity distribution for simulation

30.


4. RESULTS AND DISCUSSION

The transformation from the liquid state to the solid state is accompa-nied by a decrease in volume in most metals, leading to shrinkage and

porosity [1]. The tendency for the formation of shrinkage is associated

with both liquid and solid volume fraction and the solidification tem-perature range of the particular alloy [1, 12]. Shrinkage occurs in me-tallic materials during freezing and cooling. Using the virtual casting software, macroshrinkage and mi-croshrinkage for the 36 simulation runs were done. The contour plots

showing the location and magnitude of porosity formed during the so-lidification for simulation number 13 is shown in Fig. 4, in 2D, which

is difficult to measure. Hence, 3D solid works software was used for

generating 3D model of the images. The large tolerance level (0.1 mm),

typical in foundry processes [19] was considered for the present study.

Post processing image is converted to the solid model part (SLD-part)

file format. The starting location of the shrinkage porosity was 80 mm

from bottom and spread up to 92 mm from the flat bottom. These con-

Fig. 4. Typical contour plots showing the porosity distribution for simula-tion 13.

a b

Fig. 5. Constructed porosity of the simulation run order 13: a—84 mm from

bottom (volume is 15.49288 cm3), b—88 mm from bottom (volume is 10.8578

cm3).


tour plot data were converted to solid model part (SLD-part) was shown

as constructed porosity in Fig. 5 (for 84 mm and 88 mm). The shrinkage porosity distribution for simulation 13 between 80–92 mm from the flat bottom was given in Table 6. The porosity distri-bution obtained from the contour plot in all the four locations was

0.499. The amount of shrinkage porosity distribution in these regions

is given in Table 7. The porosity distribution for all the 36 simulations was calculated in

similar way. Figure 6 gives the constructed porosity of simulation for

run 5 and 30 for rectangle (103 mm from bottom, volume is 10.399 cm3

and 108 mm from bottom, volume is 9.4899 cm3) and cube (68 mm from

bottom, volume is 6.02598 cm3 and 72 mm from bottom, volume is

6.20088 cm3) castings respectively.

The porosity distribution results for all the 36 simulations are given

in Table 8.

4.1. Solidification

When liquid metal enters a mould cavity, its heat is absorbed and

transferred through the mould wall as shown in Fig. 7. For eutectics

and pure metals, the solidification continues layer by layer starting

from the mould wall and gradually moving inwards [21]. The progress-ing layer between the liquid and solid is called the solidification front.

TABLE 6. Shrinkage porosity distribution—Simulation 13.

Sl. No. Distance from the flat bottom, mm Volume, cm3

1 2 3 4

80 84 88 92

15.49288 10.63793 10.85780 15.55000

Total volume 52.53861

TABLE 7. Amount of Shrinkage porosity—Simulation 13.

Total volume of shrinkage porosity distribution, cm

3 52.53861

Amount of shrinkage porosity distribution in the four regions, cm

3 52.53861×0.499 = 26.21677

Amount of shrinkage porosity in the casting, cm

3 26.21677/572.265 = 0.045812

Percentage shrinkage porosity in the casting 4.5812


As this front solidifies, it contracts in volume, and draws liquid metal from the inner part of the casting (opposite to the mould wall). When the solidification front reaches the innermost region or the

thickest part of the casting, there is no liquid metal left to compensate

the shrinkage, and as a result, a void space called shrinkage cavity is

formed. When liquid metal enters into the mould cavity, its heat is removed

and transferred through the mould wall. Most of the cast alloys do not

have distinct melting point. Only pure metals and eutectic alloys solid-ify at a constant temperature. However, cast alloys solidify over a

Fig. 7. Solidification of casting in a mould.

Fig. 6. Shows constructed porosity of simulation for run: a—30, rectangle

casting; b—5, cube casting.


range of temperatures. The range between the liquidus line and solidus

line is known as the freezing range Rf = Tliq − Tsol, where Tliq is liquidus

temperature and Tsol is solidus temperature as shown in Fig. 8. Solidification proceeds layer-by-layer in short freezing range

(US 413) alloys, which behave like pure metals and eutectics. Solidifi-cation is initiated at a large number of nucleation sites, in case of long

TABLE 8. Porosity distribution for all 36 simulations.

Simulation


temperature, °C Mould coat

Casting mould

Shrinkage porosity, %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US A356 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413 US 413

MS MS MS MS MS MS MS MS MS No No No No No No No No No MS MS MS MS MS MS MS MS MS No No No No No No No No No

T + 50 T + 50 T + 50 T + 50 T + 50 T + 50

T T T

T + 50 T + 50 T + 50

T T T T T T T T T T T T

T + 50 T + 50 T + 50

T T T

T + 50 T + 50 T + 50 T + 50 T + 50 T + 50

GC GC GC NC NC NC GC GC NC NC NC GC GC GC GC NC NC NC NC NC NC GC GC GC NC NC NC GC GC GC NC NC NC GC GC GC

Cylinder Cube

Rectangle Cylinder











Cube Rectangle

3.473 2.845 2.628 3.500 3.162 2.700 3.850 3.250 2.800 3.800 3.240 2.895 4.582 3.300 3.100 4.515 3.500 3.130 3.380 2.900 2.180 3.340 2.880 2.190 3.040 2.610 2.000 3.400 2.710 2.290 3.350 2.640 2.100 3.300 2.600 2.120


freezing range alloys and grains grows until the neighbouring grains

hinder them. Freezing rate is greatly affected by the cooling rate and

thermal gradients inside the casting, which are influenced by varying

process parameters.

4.2. Influence of Process Parameters

4.2.1. Casting Mould

Progress of solidification is influenced by size, shape and thickness of

the casting [18, 20]. Macrocavities decrease with casting shape chang-es from thick cylinder to thin rectangle castings. The shape of the cast-ing influences the solidification of the casting in localized areas be-cause the edges play a significant role in the heat extraction. The edges

of the test casting, combined with the relative composition of the alloy,

decide the variation in formation of macrocavities. The solidification

time is inversely proportional to surface area of the casting [22]. Relative solidification time affects the macrocavities (with in-creased solidification time, the macrocavities increase). Solidification

time is higher in cylindrical shape castings (which have the least sur-face area of the three casting shapes considered for the present study)

[22], which promoted maximum macrocavities. The influence of cast-ing shape on shrinkage porosity is shown for the two alloys US A356

and US 413 in Figs. 9 and 10. The numbers 1–6 on the X-axis in the

Figs. 9 and 10 were percentage shrinkage values obtained for six simu-lation runs conducted for each casting mould.

4.2.2. Chill

Bottom chills reduced the shrinkage porosity for all the shapes of cast-ings (rectangle, cube and cylinder). Bottom chill extracts heat locally.

However, there was an increased rate of heat extraction from the liq-

Fig. 8. Freezing range of alloys.


uid metal and compensation of solidification shrinkage as the pouring

time of molten metal increased, giving a reduction in shrinkage porosi-ty. The rate of heat extraction and the temperature gradients inside the

solidifying metal influenced the solidification. Changes in tempera-ture gradient by the presence of chill had no effect on macrosolidifica-

Fig. 9. Influence of casting mould on shrinkage—US A356.

Fig. 10. Influence of casting mould on shrinkage—US 413.


tion and hence macrocavities were not influenced. Chills promoted

steeper temperature gradients in the solidifying metal and increased

the feeding capacity, thus reducing internal porosity and pore size.

The influence of a bottom chill on the shrinkage porosity is given in

Fig. 11.

4.2.3. Pouring Temperature

Pouring temperature is an important parameter in foundries for man-ufacturing quality castings. In order to study the influence of pouring

temperature on shrinkage porosity, pouring temperature, T, and

T + 50°C were considered. Superheat is the additional heat essential for

melting, which increases the fluidity and provides the allowance for

heat losses before they are in their final position in the mould. In-creased pouring temperature results in lower rate of heat extraction by

the mould, hence higher pouring temperature led to decreased macro-cavities and internal porosity as shown in Fig. 12.

4.2.4. Mould Coat

In order to study the effect of mould coating on shrinkage porosity,

moulds with and without graphite coatings were considered. Mould

coatings provide smooth casting surfaces and influence the thermal gradient by promoting the directional solidification. Mould coatings

Fig. 11. Influence of chill on shrinkage.


allow a passageway for feed metal to flow into the solidifying structure

and compensated for normal metal shrinkage during solidification.

There was decreased shrinkage porosity for all the three types of cast-ing shapes in graphite-coated moulds as shown in Fig. 13.

Fig. 12. Influence of pouring temperature on shrinkage.

Fig. 13. Influence of mould coat on shrinkage.


4.2.5. Interaction Plots

The importance of interactions among influencing factors are illus-trated using interaction plots. The effect of one factor is dependent

upon the other. To study interaction effects between the process pa-rameters, it is necessary to vary all the factors simultaneously. If the

process parameters are more than three, then, the interaction plot can

be displayed in the form of matrix. The plot has one panel for each pair

of factors in the data set. Parallel lines in the interaction plot imply no interaction between

the influencing factors. Increased departure from parallel plots im-plies stronger interaction effects. A significant amount of interaction was observed between the pro-cessing parameters. The interaction and main effects plot (given in

Figs. 14 and 15) for shrinkage indicate that there was a strong interac-tion among all the influencing factors, when considering the levels

(variables). Thus, the alloy, bottom chill, mould coat, casting shape

Fig. 14. Interaction plots of processing parameters on shrinkage.

Fig. 15. Plots showing the processing parameters on shrinkage.


and pouring temperature influence the shrinkage. The difference in

the vertical position of the plotted points indicates greater magnitude

of the effect, hence the bottom chill factor is considered more signifi-cant in influencing shrinkage.

5. EXPERIMENTAL VALIDATION STUDIES

Experiments were conducted to validate the casting process simulation

results. To evaluate the influence of process parameters, six experi-ments were conducted (given in Table 9), and a schematic diagram of

the testing arrangement for cube casting is shown in Fig. 16. These ex-periments were designated as experiments run order 1 to 6. The pre-pared moulds, overflow core and pouring basins for three casting

shapes are shown in Fig. 17. The overflow core was placed over the

mould in order to ensure that only a fixed quantity of metal was poured

into the mould each time.

5.1. Moulding, Melting and Pouring

Moulds were prepared using green sand process. The sand composition

TABLE 9. Conditions for validation experiments.

Experiment


temperature, °C Mould coat Casting shape

1 2 3 4 5 6

US A356 US A356 US A356 US 413 US 413 US 413

MS MS MS MS No MS

T + 50 T + 50

T T

T + 50 T

NC GC NC NC GC NC

Cylinder Cube

Rectangle Rectangle

Cube Rectangle

Fig. 16. The assembled mould for volume deficit experiment.


was silica sand with 5–6 wt.% Bentonite and 5–8 wt.% water. Moulds

were prepared with slight ramming. The patterns stripped after

3 hours. The alloy was melted in an electric resistance furnace of capac-ity 20 kg provided with mild steel crucible. Temperature was measured

with a K-type thermocouple. The furnace was turned off and the crucible was lifted and put in a

tilting device. The metal was tapped into a smaller crucible for pouring

into the mould. Figure 18 shows the solidified castings of the valida-tion experiments. Low-magnification examination of cast components reveals porosi-ty, shrinkage. Unetched condition generally discloses the porosity and

shrinkage defects. During solidification of casting, if a region in the

casting receives insufficient amount of liquid metal, then, the liquid in

this region solidifies either in the form of shrinkage/pore. Structure

consists of a network of silicon particles (sharp and grey), formed in

the interdendritic aluminium-silicon eutectic, for alloy US A356, as

sand cast is shown in Fig. 19, while the structure consists of eutectic

Fig. 17. Schematic diagram for the volume deficit experiment for cube shape

casting.

Fig. 18. Solidified castings from the validation experiments.


silicon (grey constituent) for alloy US 413, as sand cast is shown in

Fig. 20 at 50× magnification.

5.2. Shrinkage Porosity Calculations for the Validation Experiments

Shrinkage porosity calculations for the validation experiments were

done using Archimedes principle. The shrinkage porosity values for

the six validation experiments are given at Table 10.

5.3. Comparison of Simulation and Experimental Studies

To establish the strength of association between the casting process

simulation and experimental validation studies, correlation co-efficient between the two results is calculated using the formula:

( ) ( )2 22 2

correlation( ) ,N XY X Y

rN X X N Y Y

−=

− −

∑ ∑ ∑∑ ∑ ∑ ∑

Fig. 19. Alloy US A356, as sand cast.

Fig. 20. Alloy US 413, as sand cast.


where N is number of values (casting shape), X is shrinkage value of

casting process simulation, Y is shrinkage value of experimental vali-dation study. The correlation coefficient value obtained was 0.9389 indicating a

strong, positive relationship between casting process simulation and

validation experiments for US 413 alloy as shown in Fig. 21. The value of correlation coefficient for US A356 alloy was calculated

similarly and is equal to 0.9353, indicating a strong association be-tween the two studies. The simulation results are in agreement with validation experi-ments data. The casting process simulation results are in agreement with valida-tion experiments, however minor variation is observed. During solidi-fication, the mould cavity continuously changes the properties of liq-

TABLE 10. Shrinkage porosity values for the six validation experiments.

Experiment run order Alloy Chill

Pouring temperature, °C

Mould coat

Casting mould

Shrinkage porosity, %

1 2 3 4 5 6

US A356 US A356 US A356 US 413 US 413 US 413

MS MS MS MS No MS

T + 50 T + 50

T T

T + 50 T

NC GC NC NC GC NC

Cylinder Cube

Rectangle Rectangle

Cube Rectangle

2.84 2.65 2.63 2.11 2.07 1.90

Fig. 21. Shrinkage porosity: simulation and experimental studies.


uid and solid phases due to interactions between metal-mould, metal-ambient and mould-ambient. Exactly envisaging the change in proper-ties during solidification is very difficult in virtual environment (cast-ing simulations). Interfacial heat transfer coefficient (h) is one of the

critical properties of casting process simulation. Although mould and

interface are considered as one and the same, certain mechanisms oc-cur and change the h value during the casting process. It is necessary to obtain precise experimental data for mould coat, to study the effect

of heat transfer coefficient accurately. Due to the above factors, there

could be minor variation between the study results of casting process

simulation and validation experiments.

6. CONCLUSION

For casting process simulation and experimental validation studies,

the correlation coefficient value obtained is observed to be close to +1

indicating a positive relationship between them. The thermal insula-tion properties of the mould coating played a vital role in shrinkage of

the casting. The mould coating was designed to absorb the gaseous

products formed at the metal front thereby reducing the shrinkage.

Shrinkage decreased with changed casting shape, from cylinder to rec-tangle castings. Higher pouring temperature resulted in lower rates of

heat extraction by the mould, thereby decreasing macro- and mi-croshrinkage. Chill promoted faster temperature gradient in the solid-ifying metal and increase its feeding capacity thereby reducing

shrinkage.

ACKNOWLEDGEMENTS

The authors thank the Directorate of Engineering and the Director,

DRDL for providing support and permission for carrying out this R&D

work.

REFERENCES

1. ASM Metals Handbook. Vol. 15. Casting (Eds. S. Viswanathan, D. Apelian, R. DasGupta, M. Gywn, J. L. Jorstad, R. W. Monroe, T. E. Prucha, M. Sahoo, E. S. Szekeres, and D. Twarog) (ASM International: 2008).

2. Foundry Rejects Old Conventional Approach of Trial and Error and Adopts

Computer Simulation (Eds. A. Sholapurwalla, S. Scott, and C.-A. Rolle) (Casting Engineer: 2010).

3. Durgesh Joshi and B. Ravi, Indian Foundry J., 56, No. 1: 23 (2010). 4. S. Venkateswaran, R. M. Mallya, and M. R. Seshadri, AFS Trans., 94: 67

(1986). 5. Experiment Design & Analysis Reference (Tucson, Arizona, USA: 2008).


6. Design of Experiments Guide. JMP, A Business Unit of SAS, SAS Campus

Drive, Cary, NC 27513 (2010). 7. Minitab 16. User Manual of MINITAB (2008). 8. D. Hanumantha Rao, G. R. N. Tagore, and G. Rangajanardhan, Indian

Foundry J., 52, No. 5: 26 (2006). 9. S. Sundarrajan, H. Md. Roshan, and E. G. Ramachandran, Trans. Indian

Institute of Metals, 37, No. 4: (1984). 10. S. Sundarrajan, H. Md. Roshan, and E. G. Ramachandran, Proc. 37th

ATM of

IIM (Varanasi, India, 1983), p. 119. 11. S. Sundarrajan, H. Md. Roshan, and E. G. Ramachandran, Proc. IIF,

32nd Annual Conference (1983), p. 21.

12. Foseco Non-Ferrous Foundryman’s Handbook. Eleventh Edition (Ed. J. R. Brown) (Butterworth Heinemann Publisher: 1999).

13. User Manual Virtual Casting Developed by Regional Research Laboratory

(Trivandrum, India: 2006). 14. M. Trovant, A Boundary Condition Coupling Strategy for the Modeling of Metal

Casting Processes (Disser. for PhD) (Toronto, Canada: Graduate Department of

Metallurgy and Materials Science University of Toronto: 1998). 15. M. Venkataramana, V. Vasudeva Rao, R. Ramgopal Varma, and

S. Sundarrajan, J. Instrum. Soc. India, 37, No. 3: 157 (2008). 16. G. S. Cellini and L. Tomesani, J. Achiev. Mater. Manuf. Eng., 29, Iss. 1: 47

(2008). 17 J. Campbell and R. A. Harding, The Fluidity of Molten Metals (TALAT Lecture

3205) (The University of Birmingham, United Kingdom: 1994). 18. P. C. Mukherjee, Fundamentals of Metal Casting Technology (India: Oxford &

IBH Pub. Co.: 1988). 19. J. Nukaga, S. Kitazawa, and H. Kamimura, NDT & E International, 41, Iss. 7:

564 (2008). 20. L. Arnberg, Solidification Characteristics of Aluminium Alloys, Dendrite

Coherency (AFS: 1996), vol. 3. 21. R. F. Mehl, Proc. of Symp. ‘The Solidification of Metals and Alloys’ (Feb. 12,

1951) (New York, USA: AIME: 1951), p. 24. 22. R. W. Heine, C. R. Loper Jr., and P. C. Rosenthal, Principles of Metal Casting

(New York, USA: McGraw-Hill: 1967).

https://doi.org/10.1016/j.ndteint.2008.04.003

https://doi.org/10.1016/j.ndteint.2008.04.003

Samavedam Santhi and Srinivasan Sundarrajan

Documents

Transcript of Samavedam Santhi and Srinivasan Sundarrajan